Mathematical Modeling of Pneumatic Char Injection in a Direct Reduction Rotary Kiln

V. RAMAKRISHNAN and P.S.T. SAI A one-dimensional steady-state mathematical model is proposed for direct reduction process in rotary kilns akin to the SL/RN process. The model takes into account pneumatic coal char injection from the discharge end of the kiln to supplement the heat availability. The model is based on material and energy conservation principles, and the empirical equations for kinetics and heat transfer are obtained from the literature. Predictions are carried out for both iron oxide reduction and ilmenite beneficiation processes. Improvement in the performance was predicted with pneumatic char injection.

I. INTRODUCTION

DIRECT reduction processes are widely known alternatives for the blast furnace route to iron manufacture. Direct reduction processes in rotary kilns are also used for the beneficiation of low-quality iron ores and ilmenite. A bibliographic survey of the direct reduction processes for iron ore reduction has been complied by The Metals Society (London),[1] which gives information on the various processes that are of academic and industrial significance. In direct reduction rotary kilns, the solid reductant serves the purpose of both fuel and reducing agent. Heat is supplied to the charge by the hot freeboard gas flowing countercurrent to the solids through convection, radiation, and also regeneratively through the walls. Heat is generated in the freeboard by the combustion of CO released from the bed. The desired temperature is attained by the controlled addition of the combustion air along the length of the kiln. A number of fans are installed on the shell of the kiln, which supply the necessary process air via the so-called air tubes or via air injecting nozzles in the feed zone of the rotary kiln. Direct reduction processes for iron ore reduction in rotary kilns have been modeled by Wingfield et al.,[2] Venkateswaran and Brimacombe,[3] and Mukhopadhyay et al.[4] Of these models, the former is for gaseous reductant, while the other two are for coal-based direct reduction process. The model proposed by Venkateswaran and Brimacombe[3] is a comprehensive and yet simple model. In spite of the simplifying assumptions made in the kinetics, dispersion, and heat transfer, the model calls for the solution of a boundary value problem of 12 first-order differential equations. More rigorous modeling work on the direct reduction process has not been reported. Many mathematical models have been proposed for operations in rotary kilns with heat transfer to the solids bed. Gorog et al.[5] studied the radiative heat transfer between a nongray freeboard gas and the interior surfaces of a rotary kiln by evaluating the fundamental radiative exchange integrals to develop a model for the net heat flux to the solids
V. RAMAKRISHNAN, Graduate Student, formerly with the Department of Chemical Engineering, Indian Institute of Technology, is with the Department of Chemical Engineering, University of Connecticut, Storrs, CT 06269-3222. P.S.T. SAI, Associate Professor, is with the Department of Chemical Engineering, Indian Institute of Technology, Madras-600 036, India. Manuscript submitted July 9, 1998.
METALLURGICAL AND MATERIALS TRANSACTIONS B

and to the kiln and compared to a simple resistive network model. They also developed models to determine the temperature distribution in the wall of a rotary kiln[6] and heat flow in the flame zone of a direct-fired rotary kiln.[7] Barr et al.[8,9] proposed a model for rotary kiln heat transfer that accounts for the interaction of all the transport paths and processes. They applied a ray tracing technique to derive coefficients for radiative heat transfer in the kiln freeboard and extended the finite difference model into the contacting bed material in order to calculate the exchange between the covered wall and the bed. Pneumatic char (or coal fines in some cases) injection is being carried out in the present day direct reduction operations in rotary kiln.[10] This generally serves two purposes. One, in providing heat to the freeboard gas by oxidation and, two, in increasing the carbon content in the bed and maintaining a reducing atmosphere in the bed. Coal/char injection rate, char particle size, and angle of injection pipe are the parameters that characterize the direct reduction process. A commercial process for ilmenite beneficiation using direct reduction in rotary kilns in Western Australia[10] uses pneumatic coal injection to supply the additional heat required in the reduction zone. Coal char injection involves the storage of particles in an unsealed silo continuously supplied with material. The particles are released from the silo by a rotary air lock feeder into a high velocity air stream of the injector system. Air from a supply source is allowed to expand to the system pressure through a variable nozzle to achieve desired flow rates. The particles are carried through the ducting by the air stream to a suitable outlet in the kiln where it is let out as a free jet. The particles reach the bed tracing an approximate path of a projectile falling under the influence of gravity. Pneumatic injection of the coal/char particles is achieved by utilizing the kinetic energy or flow momentum of air used for injection, using either more air at low velocities or less air at high velocities. The dynamics of the injected particles and air flow requirements are reported by Biswas,[11] who deals with the design requirements of the coal injector for the coal-based direct reduction process in rotary kilns. No published literature, which will predict the influence of coal char injection on the performance, is available on the modeling of the process. A number of mathematical models for combustion of coal char particles have been developed, vary depending upon
VOLUME 30B, OCTOBER 1999969

C (1/2)O2 CO

[2]

The reaction rates for the two char-gas reactions as reported by Caram and Amundson[13] are used in the present work and they are For Char-O2 reaction, R1 k1 C1s For the Char-CO2 reaction, R2 k2 C2s
Fig. 1Coal char particle undergoing combustion.

[3]

[4]

where k1 3.007 105 e(17,966/Tp) [5] [6] and k2 4.1 109 e(29,790/Tp) The reaction rate at the surface equals the mass-transfer rate and for the two reacting gases R1
R2

the simplifying assumptions used in each case. The models of significance are the ones developed by Yagi and Kunii,[12] Caram and Amundson,[13] and Essenhigh.[14] Caram and Amundson have compared various models and have discussed the shortcomings of each. In the present study, an attempt is made to develop a simple and yet comprehensive model of direct reduction process with pneumatic injection of char so as to retain the mathematical tractability. A mathematical model for char combustion inside the kiln is formulated based on shrinking core principle and is integrated with the direct reduction process along with the particle dynamic equations. Thus, the model takes into account the influence of kinetics, heat transfer, and operating variables such as air distribution and pneumatic char injection on the performance of the process.
II. MODELING OF COAL CHAR COMBUSTION As proposed here, to integrate the mathematical model for coal char combustion with the model of the direct reduction process, certain simplifying assumptions are made, which makes it adaptable within reasonable margins of error. They are as follows:
(1) Pneumatically injected char particles are divided into different size ranges and particles of each size range are lumped. (2) The particles are assumed to be spherical and are characterized by the mean size (diameter) of the size range. (3) Both O2 and CO2 react at the char surface to give CO. (4) The freeboard gas in the kiln around the particle as a whole is lumped (across the cross section), and the volume of the gas film is assumed to be insignificant in comparison to the bulk volume. (5) The combustion of the CO is assumed to occur in the freeboard gas. This assumption is reasonable considering the fact that the gas film is assumed to be small, and at high temperatures at which the reaction takes place, CO is generally observed to be the main product at low O2 partial pressures. (6) The resistance due to the ash layer that forms on the char particle is assumed to be very small considering the particle velocities in the kiln. (7) The products of the reaction freely diffuse out of the gas film surrounding the char particle.
Reactions that occur at the surface of a pneumatically injected char particle shown in Figure 1 are C CO2 2CO
970VOLUME 30B, OCTOBER 1999

k1km1 C 0.5k1 km1 1

k2km2 C k2 km2 2

[7]
[8]

The mass-transfer coefficient through the boundary film around the carbon particle, km , can be calculated using the expression given by Ranz and Marshall.[15] NNu km dp f 2.0 0.6(NSc)1/3(NRem)1/2 Dvf
ddp 2 (R1 R2)Nc dt P
The heat balance equation for the particle is dHP 4 4 d2 P {hp(TG TP) [GG(T G T P) dt
4 4 4 SP(T 4 S T P) WP(T W T P)]}

[9]

The mass conservation equation for the particle is

[10]

[11]

On simplification, the preceding equation reduces to dTP 6 4 {h (T TP) [GG(T 4 G T P) dt pdPCP p G

4 4 4 SP(T 4 S T P) WP(T W T P)]

[12]

(HB)R1 (HR)R2} The convective heat-transfer coefficient for the heat transfer between the coal char particle and the freeboard gas, hp , can be calculated by following the heat-transfer analogous of Eq. [9], as reported by Ranz and Marshall: Nu hpdp 2.0 0.6 (NPr)1/3 (NRem)1/2 k( f )
[13]

[1]

The preceding equations are coupled with the particle dynamic equations given by Biswas,[11] as presented in the Appendix, and are used in the model of the direct reduction process. The material and energy balance equations for the solid and gas phases as proposed by Venkateswaran and Brimacombe[3] are described in the Appendix. The velocity
METALLURGICAL AND MATERIALS TRANSACTIONS B

of the conveying air is determined from the particle velocity, based upon the particle size and host of other factors. Biswas[11] has tabulated the expressions for the air conveying velocities for different particle sizes. The solution method employed in the present study is detailed subsequently. (1) When integrating the differential equations for the direct reduction process from the discharge end, at each point along the length of the kiln for a given increment in the independent variable (distance from the discharge end), the heat- and mass-transfer rates per unit volume of the kiln are determined; (2) These are obtained by determining the heat- and masstransfer rates (for gases O2, CO2, and CO) for a single particle in a given size range and multiplying by the number of particles per unit volume of the kiln for the size range, Np , given as follows: NP 6fiW 1 3 2pr 2 d 1 P uh
[14]

(3) Once the particle reaches the solids bed, or the particle size reduces to less than 5 pct of its original size, the particles are no longer assumed to undergo combustion. In the former case, they are assumed to get mixed with the bed and no corrections are made to the carbon concentration in the bed as the amount injected is very small in comparison to the amount of coal in the bed. In the latter case, the particle is assumed to be carried away with the gases. This point varies for the different size ranges. The integration is then carried out as detailed subsequently.

III. BOUNDARY CONDITIONS AND NUMERICAL SOLUTIONS The model comprised of coupled differential and algebraic equations has to be numerically solved to obtain the prediction for a given set of conditions. The boundary conditions are defined for the solid charge at the feed end and the freeboard gas at the discharge end. The initial condition for the coal particles and the conveying air are defined at the discharge end. This is a two-point boundary problem of firstorder differential equations. The solution uses the fourthorder RungeKutta method for the solution of the differential equations, and shooting to a fitting point approach was used, wherein initial conditions are assumed at either ends, and the variables are compared at a certain point halfway. The error, i.e., sum of square of the normalized differences between the variables, at this point was minimized by changing the initial conditions using the globally convergent NewtonRaphson method. The flow chart is shown in Figure 2. The accepted level of error was fixed as 0.1 pct. This approach of shooting to a fitting point is essential due to the instability encountered when proceeding from the charge end, while, when proceeding from the discharge end, one is left out with only one initial condition to match, and hence by proceeding to a point halfway from either end and matching the variables, these problems can be avoided. But one encounters a problem of solving simultaneous equations with seven unknowns and computation of the Jacobian at each step to choose the direction and step to minimize the error. As alternative
METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 2Flow chart for the solution of the mathematical model.

approaches are not better, the previously described method is used. IV. RESULTS AND DISCUSSION The prediction of the mathematical model proposed here is compared with the experimental data for the ilmenite beneficiation process carried out in rotary kilns. The solid bed temperature and reduction profiles of ilmenite beneficiation carried out at The Western Titanium Limited (Capel, Western Australia) were obtained from Hockin.[10] The beneficiation process is based on the direct reduction of iron oxide content of the ilmenite to metallic iron in a rotary kiln using a highly volatile coal as both fuel and reductant. The conditions that were used for the runs for this system are shown in Table I. Other relevant information such as the kinetic expressions are given in the Appendix. Owing to lack of information available on the amounts of coal char injected and the air distribution used, the predictions have been made by assuming the amount of coal injected and choosing an air distribution shown in Figure 3 such that the predictions reasonably agree with the experimental observations. Though one cannot claim this as a proper confirmation,
VOLUME 30B, OCTOBER 1999971

Table I. Input Parameters for the Ilmenite Beneficiation Process

Kiln characteristics Length of the kiln, m Degree of fill, pct Outer shell diameter, m Inner shell diameter, m Inner diameter of outer refractory, m Inner diameter of inner refractory, m Input characteristics Ilmenite feed rate, tons/annum Coal feed rate, tons/annum Pneumatic coal injection rate, tons/annum Initial particle velocity, m/s Solid to air ratio Composition of charge, pct Preoxidized Ilmenite TiO2 53.5 FeO 0.6 Fe2O3 44.7 MnO 1.3 Coal Fixed carbon Moisture Volatiles Ash Sulfur 45 18 3.5 3.4 3.16 2.59 53,225 19,054 1710 63.2 3.25

42.1 25.0 27.4 4.8 0.7

Size distribution of pneumatically injected coal char particles, pct 12.70 mm 0.0 12.70 mm 6.35 mm 4.5 6.35 mm 3.34 mm 43.0 3.34 mm 1.40 mm 21.0 1.40 mm 1.00 mm 27.5 1.00 mm 4.0

Fig. 4Comparison of experimental and predicted solids bed temperature and degree of reduction of ilmenite.

Fig. 5Temperature distribution for direct reduction of ilmenite.

Fig. 3Air distribution used for the ilmenite beneficiation process.

it definitely indicates that the model predictions may not be off by a large margin, and the model will be helpful in giving at least a rough idea of the performance where no results exist. The comparisons are shown in Figure 4. The variation of the solid, gas, wall, and shell temperatures along the length of the kiln is shown in Figure 5, and the
972VOLUME 30B, OCTOBER 1999

variation of partial pressure and degree of reduction are shown in Figure 6. The general trends are the same as the ones one would expect in a direct reduction process. In order to make some comparisons regarding the influence of coal injection, a run was carried out without coal char injection. The comparison of solids bed temperature and the degree of reduction is shown in Figure 7. It is seen that the lower heat availability results in lower solid temperature and hence lower degree of reduction. In order to show the ease of adaptability of the model for different systems, the influence of coal char injection is also studied for iron oxide system. For the present study, the same process conditions as used by Venkateswaran and Brimacombe[3] for iron oxide reduction are considered to be
METALLURGICAL AND MATERIALS TRANSACTIONS B

Table II. Input Parameters for the Iron Oxide Reduction Process
Kiln characteristics Kiln length, m Inner radius of the kiln, m Outer radius of inner refractory, m Outer radius of outer refractory, m Outer shell radius, m Thermal conductivity of the inner refractory, cal cm1 s1 K1 Thermal conductivity of the outer refractory, cal cm1 s1 K1 Input characteristics Ore feed rate, tons/day Coal feed rate, tons/day Air feed rate, m3/ton ore Natural gas rate, m3/ton ore Inlet charge temperature, K composition, pct
Ore
Fig. 6Partial pressure and reduction profiles for direct reduction of ilmenite.

30 1.05 1.28 1.33 1.36 0.0031 0.0005

174 118 1645.3 21.2 298

Coal Fixed carbon Moisture Volatiles Ash Sulfur content

Hematite Moisture Gangue

95.5 1.4 3.1

42.0 31.0 16.7 10.0 0.3

Iron concentration in bed, g/cm3 Carbon concentration in bed, g/cm3 Oxide pellet reducibility, cm3 g1 s1 Coal reactivity, cm3 g1 s1

0.763 0.275 3.143 103 51.9 1016

Table III. Air Distribution Used for the Iron Oxide Reduction Process
Air rate, m3/ms Distance from the Charge End, m 0 to 4 4 to 19 19 to 26 26 to 30 0.1 0.14 0.10 0.00

Fig. 7Comparison of solids bed temperature and reduction profiles for direct reduction of ilmenite with and without char injection.

a meaningful comparison. The conditions that are common for all studies carried out with this system are shown in Table II. The air profile used here is given in Table III. A run was carried out without natural gas or coal char injection and subsequently a run was carried out with coal injection, and the data used in relation to coal injection were as follows: solid to air ratio 1.0, particle velocity 45 m/s, rate of char injection 0.174 kg/s, and primary air from discharge end 0.1467 m3/s. The comparison of solid and gas temperatures, degree of reduction, and oxygen partial pressure for the two cases is
METALLURGICAL AND MATERIALS TRANSACTIONS B

shown in Figure 8 through 10, respectively. As for the ilmenite reduction predictions, it was seen that the solid and gas temperatures were higher and hence a higher degree of reduction was predicted. It was also predicted that with coal char injection, the ratio of the reduction to preheat zone length was increased, which would be an indication of the better utilization of the rotary kiln. The variation in oxygen partial pressure indicated that more air than required was present in the case without coal char injection and hence dilution of temperature and poor performance resulted. V. CONCLUSIONS A mathematical model for the direct reduction process with coal char injection has been proposed, and coal char injection predicted improved performance of the direct reduction process. The uncertainty involved in the range of pneumatically injected particles had a bearing on the predictions. It can be improved by determining the average range empirically for given kiln geometry and particle size.
VOLUME 30B, OCTOBER 1999973

Fig. 8Comparison of solids bed and freeboard gas temperature distribution for direct reduction of iron oxide with and without char injection. Fig. 10Comparison of oxygen partial pressure profile for direct reduction of iron oxide with and without char injection.

Dvf EB , ER fi F g G Ggi, Gsi hp hcws hgsc hgsr hgwc

Fig. 9Comparison of reduction profile for direct reduction of iron oxide with and without char injection.

hgwr hsh hssc hssr hwsr H HC Hp k( f ) k1, k2 ka

NOTATION a A A1, A2, A3, A4 Ap C1, C2 C1s, C2s Cp dp D acceleration, m/s2 cross-sectional area of the kiln, m2 areas as shown in Fig A1, m2 reducibility of ore, cm3/gs concentrations of O2 and CO2 in the bulk, respectively, kmol/m3 concentrations of O2 and CO2 at the surface of the char particle, respectively, kmol/m3 Specific heat capacity, kcal/kgK diameter of the char particle, m diameter of the kiln, m

diffusivity of O2 in the freeboard gas, m2/s activation energies for Boudouard and reduction reactions, respectively, cal/gmol weight fraction of the ith size range degree of reduction acceleration of gravity, m/s2 mass flow rate of the freeboard gas, kg/s mass flux of the ith component in the gas and solid phases, respectively, kg/m2s heat-transfer coefficient between coal char particles and the freeboard gas, kcal/m2sK heat transfer coefficient for covered wall to solids (across area A3), kcal/ms2sK heat-transfer coefficient for convection from gas to solid (across area A2), kcal/m2sK heat-transfer coefficient for radiation from gas to solid (across area A2), kcal/m2sK heat-transfer coefficient for convection from gas to wall (across area A1), kcal/m2sK heat-transfer coefficient for radiation from gas to wall (across area A1), kcal/ms2sK heat-transfer coefficient for heat loss from wall to ambience (across area A4), kcal/m2sK heat-transfer coefficient for convection loss from shell, kcal/m2sK heat-transfer coefficient for radiation loss from shell, kcal/m2sK heat-transfer coefficient from wall to exposed solids (across area A2), kcal/m2sK enthalpy flux, kcal/m2s reactivity of coal, cm3/gs enthalpy of the particle, kcal average thermal conductivity of the freeboard gas, W/mK reaction rate constants for char-O2 and char-CO2 reactions, respectively, m/s thermal conductivity of ambient air, kcal/msK
METALLURGICAL AND MATERIALS TRANSACTIONS B

974VOLUME 30B, OCTOBER 1999

kI , k O KB , KR km1, km2
MC , MFe N NC NGr NNu NP NPr NRem NRew NSc PB CO PCO2 PB CO2 PR CO2 Qwg Qwl Qws qcg, qcs qgg, qgs rgi, rci r1 r2, r3 r4 R R1, R2
RB , RR S T u ur
va vx x xd W

thermal conductivities of inner and outer refractory in kiln lining respectively, kcal/msK equilibrium constants for Boudouard and reduction reactions, respectively mass-transfer coefficients of O2 and CO2 in the gas film, respectively, m/s concentrations of carbon and iron in the charge, respectively, g/cm3 charge rotational speed of the kiln, s1 molecular weight of char particle grashof number ( D32g(Tsh TG)/2) nusselt number number of particle per unit volume of the kiln prandtl number ( Cp/ka) modified Reynolds number ( uhpdP /) rotational Reynolds number ( ND2/) schmidt number ( v/Dvf ) equilibrium partial pressure of CO for Boudouard reaction, atm bulk partial pressure of CO2 in the bed, atm equilibrium partial pressure of CO2 for Boudouard reaction, atm equilibrium partial pressure of CO2 for reduction reaction, atm heat-transfer rate from wall to gas, kcal/s heat loss from wall to ambience, kcal/s heat-transfer rate from wall to solids, kcal/s rate of consumption of heat in gas and solid phases, respectively, kcal/m3s rate of generation of heat in gas and solid phases, respectively, kcal/m3s rate of generation and consumption of the component i, respectively, kg/m3s inner radius of the kiln, m outer radius of the inner and outer refractory, respectively, m outer radius of the kiln shell, m gas constant, 1.987 cal/molK rate of char-O2 and char-CO2 reactions, respectively, kmol/m2s rate of Boudouard and reduction reactions, respectively, kmol/m2s displacement, m temperature, K velocity of the char particle, m/s relative velocity between the particle and the air, m/s velocity of air for pneumatic char injection, m/s velocity at a horizontal location x, m/s distance from charge end of the kiln, m distance from discharge end of the kiln, m rate of char injection, kg/s

G v

standard free energy change, cal/mol viscosity of air, kg/ms kinematic viscosity of air, m2/s

Subscripts a ambient air B boudouard reaction G freeboard gas h horizontal p char particle R reduction reaction S solids bed sh shell t at time t t t at time t t t, t t between t and t t time interval v vertical W kiln wall

APPENDIX A. Material and Energy Balances In a direct reduction of a metal oxide with carbon in a rotary kiln, the reactions are given by the following equations: MexOy (s) CO (g) MexOy1 (s) CO2 (g) [A1] CO2 (g) C (s) 2CO (g) [A2]

The material and energy balances for solid and gas phases over an infinitesimally small cylindrical element along the length of the kiln at a distance x from the charge end are as follows. 1. Material balance equations For solids, dGsi rgi rci 0 dx
[A3]

where i denotes the components such as coal and ore. For gases, dGgj rgj rcj 0 dx
[A4]

Here, j stands for gases such as CO, CO2, O2, N2, H2, and CH4. 2. Heat balance equations For solids, dHS qgs qcs 0 dx
For gases, dHg qgg qcg 0 dx
For the kiln wall, Qwg Qws Qwl 0 [A7] [A6]

Greek Symbols G absorptivity of the freeboard gas coefficient of thermal expansion of air, K1 decceleration rate, m/s2 emissivity f volume fraction of the gas film around the char particle density, kg/m3 stephanBoltzmann constant, kcal/m2sK4 H standard enthalpy change, cal/mol
METALLURGICAL AND MATERIALS TRANSACTIONS B

[A5]

VOLUME 30B, OCTOBER 1999975

0.9 and 1.0, respectively. The emissivity and absorptivity data for different concentrations of CO2 and H2O are obtained from Perry and Green,[16] and a regression fit is used to calculate the emissivity and absorptivity of the freeboard gas. All the physical properties of the gases are computed by taking the weighted average using the mole fraction of the components. The experimental data reported by Venkateswaran and Brimacombe[3] for drying and volatilization of coal are fitted to a polynomial expression and are used. B. Kinetic Expressions Used for the Ilmenite Beneficiation Process The reaction scheme is C CO2 2CO FeTiO3 CO FeTiO2 CO2 The rate expression for the Boudouard reaction is given by RB MCHC eEB/RTS PCO2 PB CO2
RTS PR CO2 PCO2
RTS

[A17]

and that for the reduction reaction is given by RR MFe AP F (1 F ) eER/RTS

Fig. A1Cross section of the kiln showing heat flow processes and the corresponding areas (Venkateswaran and Brimacombe[3]).

[A18]

The equilibrium partial pressure of CO2 for the Boudouard reaction and the reduction reaction can be evaluated using KB eGB/RTS e(40,800/RTS41.7/R)
2 (1 PCO2) (PB CO) B PCO2 PB CO2
B 2

To determine the heat fluxes at different positions in the kiln, the kiln internal surface is divided into various zones, as shown by the cross section of the kiln in Figure A1. The heat-transfer coefficients for various modes of heat transfer between solids, gases, and kiln are as follows:
hgsr

[A19]

(gT 4 g

(Tg TS)

4 gT S)

KR eGR/RTS e(2890/RTS7.82/R)
PB CO2
PB CO PB CO2
1 PB CO2

[A8]
[A9]
[A10] [A11]
[A12] [A13]

[A20]

hgsc 2.63(103)(G/A)0.067 hgwr

4 (gT 4 g T W) g (T g T W )

C. Kinetic Expressions Used for the Iron Oxide Reduction Process

The reaction scheme is C CO2 2CO
Fe2O3 3CO 2Fe 3CO2 The empirical first-order rate expression for the reduction reaction as proposed by Von Bogdandy and Engell[17] is as follows:
RR 4.48(102) eER/RTSAP(1 F )MFe

hgwc 2.63(103)(G/A)0.067 hwsr

4 WS(T 4 W T S) (TW TS)

hcws 5hgsc hsh

1 r 2 r1 2 r4 k1(r1 r2)

r 3 r2 1 k0(r3 r2) 2 r4(hssc hssr)

[A14]

PR CO2 PCO2
PR CO2

[A21] The rate expression for the Boudouard reaction is the same as that for the ilmenite beneficiation process (Eq. [A17]).
[A15] D. Particle Dynamic Equations
[A16] The air mixed with coal particles upon leaving the outlet of the injector is assumed to expand as a turbulent free jet. The longitudinal distribution of velocity along the central
METALLURGICAL AND MATERIALS TRANSACTIONS B

where hssc D 0.35 0.11[(0.5N 2 Rew NGr) NPr] ka

4 sh(T 4 sh T a) hssr Tsh Ta

The emissivities of the kiln wall and solid bed are taken as
976VOLUME 30B, OCTOBER 1999

line in the stabilized zone obtained from Perry and Green[16] is given as follows:
dp vx 2.28 va xd

REFERENCES
1. Direct Reduction of Iron OreA Bibliographic Survey, The Metals Society, London, 1979. 2. S.L. Wingfield, A. Prothero, and W. Karvecki: J. Inst. Fuel, 1974, vol. 47, pp. 64-72. 3. V. Venkateswaran and J.K. Brimacombe: Metall. Trans. B, 1977, vol. 8B, pp 389-98. 4. P.K. Mukhopadhyay, A.V. Sathe, and Amit Chatterjee: Trans. Ind. Inst. Met., 1984, vol. 37 (6), pp. 721-28. 5. J.P. Gorog, J.K. Brimacombe, and T.N. Adams: Metall. Trans. B, 1981, vol. 12B, pp. 55-70. 6. J.P. Gorog, T.N. Adams, and J.K. Brimacombe: Metall. Trans. B, 1982, vol. 13B, pp. 153-63. 7. J.P. Gorog, T.N. Adams, and J.K. Brimacombe: Metall. Trans. B, 1983, vol. 14B, pp. 411-24. 8. P.V. Barr, J.K. Brimacombe, and A.P. Watkinson: Metall. Trans. B, 1989, vol. 20B, 391-402. 9. P.V. Barr, J.K. Brimacombe, and A.P. Watkinson: Metall. Trans. B, 1989, vol. 20B, 403-19. 10. H.W. Hockin: Proc. 101st Annual Meeting AIME, TMS, Warrendale, PA, 1972, pp. 209-31. 11. D.K. Biswas: Powder Handling Processing, 1993, vol. 5 (2), pp. 145-52. 12. S. Yagi and D. Kunii: Proc. 5th Int. Symp. on Combustion, Reinhold, New York, NY, 1955, pp. 231-39. 13. H.S. Caram and N.R. Amundson: Ind. Eng. Chem. Fund., 1977, vol. 16 (2), pp. 171-81. 14. R.H. Essenhigh: in Coal Conversion Technology, C.Y. Wen and E. Stanley Ley, eds., Addison-Wesley Publishing Company, Inc., Reading, MA, 1979, pp. 171-304. 15. W.R. Ranz and W.R. Marshall, Jr.: Chem. Eng. Prog., 1977, vol. 48, pp. 141-46. 16. R.H. Perry and Don Green: Perrys Chemical Engineering Handbook, 6th ed., McGraw-Hill Publications, New York, NY, 1984. 17. L. Von Bogdandy and H.J. Engell: The Reduction of Iron Ores, SpingerVerlag, Berlin, 1971, pp. 286-310.

0.5

for 5

xd 2000 dp

[A22]

The horizontal and vertical components of the velocity and displacement of the particles can be computed using the following equations: [uh]tt [uh]t [ah]t,tt t [uv]tt [uv]t [av]t, tt t [A23] [A24]

[Sh]tt [Sh]t 0.5 ([uh]t [uh]tt) t [A25] [Sv]tt [Sv]t 0.5 ([uv]t [uv]tt) t [A26] where [ah]t, tt [ur]2 t 2 gdp
[uv]2 t 2 gdp [A27]
[A28] [A29] [A30]

[av]t, tt [av]tt, t
[ur]t [uh]t [vh]t [vh]t va [vh]t 2.28va

for [Sh]t 15 dp

dp [Sh]t

0.5

for [Sh]t 15 dp

0.137 for dp 3.00 mm 0.220 for dp 1.00 mm 0.270 for dp 1.00 mm [A31]

METALLURGICAL AND MATERIALS TRANSACTIONS B

VOLUME 30B, OCTOBER 1999977